A Dynamic Prediction Model for Renal Progression in Primary Membranous Nephropathy

Area of Science:

• Nephrology
• Medical Informatics
• Predictive Modeling

Background:

• Primary membranous nephropathy (PMN) is a leading cause of nephrotic syndrome in adults.
• Predicting renal progression in PMN is crucial for timely intervention and management.
• Existing prediction tools may lack dynamic capabilities or real-world validation.

Purpose of the Study:

• To develop and validate a practical, web-based dynamic prediction model for renal progression in PMN patients.
• To identify key clinical and serological factors associated with adverse renal outcomes in PMN.
• To provide a user-friendly tool for clinicians to assess individual patient risk.

Main Methods:

• A cohort of 461 PMN patients was divided into derivation (n=359) and validation (n=102) groups.
• Renal progression was defined as a ≥30% eGFR decline or End-Stage Renal Disease (ESRD).
• Multivariable Cox regression identified predictors; a dynamic web-based model was constructed and validated using ROC and decision curve analysis.

Main Results:

• The final model incorporated hyperuricemia, proteinuria, serum albumin, eGFR, age, and sPLA2R-ab levels.
• The web-based model demonstrated good discrimination (C-statistic = 0.72) and calibration in the validation cohort.
• Key predictors included lower serum albumin, lower eGFR, higher proteinuria, hyperuricemia, older age, and higher sPLA2R-ab levels.

Conclusions:

• A validated web-based dynamic prediction model for renal progression in PMN has been developed.
• This tool can assist clinicians in identifying high-risk PMN patients.
• The model facilitates personalized treatment and surveillance strategies for improved patient outcomes.